Context (i.e., background cues) has long been known to play a vital role in modulating acquired behavior. During the past 20 years, researchers have extensively studied the effects of external context upon the extinction of Pavlovian conditioning. Recent advances in this line of research, which have direct implications for how we conduct cue exposures treatment, have yet to be incorporated into cue exposure research and treatment for alcoholics. The present application is designed to fill that void by investigating whether a recent theoretical model of extinction performance, which has been supported in laboratory animal studies, will apply to the extinction of alcohol cue responses in alcoholics. Specifically, the proposed study represents an initial test of the effect of context on measures of craving and alcohol cue reactivity in treating-seeking alcoholics. Subjects will be 210 alcoholics seeking treatment and randomly assigned to one of six experimental groups that vary with regard to the post-extinction assessment context. Using a counter-balanced design, Groups 1 through 4 will be compared to test if post-extinction assessment context affects renewal of reactivity to alcohol beverage cues. Groups 5 and 6 are included to test if the presence of a cue (i.e, extinction cue) that was paired with the extinction context can reduce craving and alcohol cue reactivity during the post-test and whether a procedure designed to increase the saliency of this cue (i.e., Group 6) will further enhance this effect. Groups 5 and 6 will be compared to Group 2 from the four group design. The short-term goal of this project is to apply recent findings from contemporary learning theory to research aimed at reducing craving and alcohol cue reactivity in alcoholics. The long-term goal is to develop alternative methods of conducting cue exposure treatment that will enhance the generalization of cue exposure treatment effects and improve treatment outcomes by further reducing alcohol cue reactivity and increasing the latency of return to drinking and heavy drinking. In summary, this study represents the first step in a programmatic line of research in which a theory developed with basic laboratory research is first tested in a human clinical sample, and then used to develop a new behavioral intervention.